Trans. JSASS Aerospace Tech. JapanVol. 12, No. ists29, pp. Pk_7-Pk_12, 2014

Original Paper

Pk_7

Planetary Atmosphere Wind Tunnel Tests on Aerodynamic Characteristics of a Mars Airplane Scale Model

By Masayuki ANYOJI1), Masato OKAMOTO2), Hidenori HIDAKA2), Taku NONOMURA1),Akira OYAMA1) and Kozo FUJII1)

1)Institute of Space and Astronautical Science, Japan Aerospace Exploration Agency, Sagamihara, Japan2)Department of Aeronautics, Kanazawa Institute of Technology, Nonoichi, Japan

Received June 22nd, 2013

Low-density wind tunnel tests are carried out using a 1/5-scale model to evaluate the aerodynamic performance of a Mars airplane at low Reynolds numbers (Re = 10,000-33,000). Particularly, Reynolds number dependency on the aerodynamic characteristics and effects of the elevator are investigated. There is little change of the lift curve and the pitching moment in the Reynolds number range from 10,000 to 33,000. The effectiveness of the elevator is not enough for responsive pitch control. It is found that the low effectiveness of the elevator is due to a non-linearity of the lift curve of the horizontal-tail airfoil and a thinner tail-airfoil having a large lift-curve slope can improve the pitch control.

Key Words: Mars Airplane, Low Reynolds Number, Aerodynamic Performance, Wind Tunnel Test

Nomenclature

ac : aerodynamic center AoA : angle of attack [deg] AR : aspect ratio c : chord length

DC : drag coefficient cg : center of gravity

LC : lift coefficient mC : pitching moment coefficient

D : drag [N] L : lift [N]

tl : distance from cg to ac of tail wl : distance from cg to ac of main wing

DL / : lift to drag ratio eR : Reynolds number

e : elevator angle [deg]

1. Introduction

A Mars airplane with a fixed wing has been considered as one of promising new exploration methods.1-4) In Martian atmospheric flight, the Reynolds number based on the chord length of the main wing becomes approximately 104 to 105

due in large part to the low density of the Martian atmosphere. In such a low Reynolds number regime, the flow around the airfoil are vastly different from those in high Reynolds numbers due to the viscous effects, and it is known that the maximum lift-to-drag ratio of smooth airfoils significantly deteriorates.5) Therefore, the aerodynamic design of the main wing is an important research issue to realize a Martian atmospheric flight.

A great deal of research on airfoil characteristics in the

low Reynolds number region has been reported in the past studies.6-9) Schmitz10, 11) has suggested three shape characteristics leading to high aerodynamic performance in a low Reynolds number region: 1) a sharp leading edge which fixes the separation point at the edge and can improve its Reynolds number dependence on the aerodynamic performance; 2) a flat upper surface which reduces the separation region; 3) a cambered airfoil which gains a higher lift than a symmetric airfoil. An airfoil named “Ishii airfoil” is one of the airfoils having above features. Mr. Mitsuru Ishii who was a world champion of a free flight contest of hand launch gliders originally designed Ishii airfoil for a main wing of a hand launch glider. In the previous experimental and numerical studies of Ishii airfoil,12, 13) it has been clarified that Ishii airfoil has a high lift-to-drag ratio (L/Dmax is about 18) and little Reynolds number dependence on its aerodynamic performance at Re = 23,000. Oyama14) and his group designed a Mars exploration airplane with Ishii airfoil as the main wing using the multiobjected design exploration methodology.

In this study, to investigate aerodynamic characteristics of the Mars airplane that was designed by Oyama and his group, low-density wind tunnel tests were conducted in the Reynolds number range from 10,000 to 33,000 using a 1/5-scale model. These tests were carried out using a low-density wind tunnel, “Planetary Atmosphere Wind Tunnel” at JAXA (Japan Aerospace Exploration Agency) that enables aerodynamic measurements at low Reynolds numbers down to Re = 8.8 x 103 (characteristics length = 100 mm) by depressurizing the wind tunnel. We focused on the Reynolds number effects on basic aerodynamic performance and the elevator effectiveness for the pitch control. Besides, the validation study was also conducted, compared with the results using a low-speed wind tunnel at Kanazawa Institute of Technology (KIT).

Copyright© 2014 by the Japan Society for Aeronautical and Space Sciences and ISTS. All rights reserved.

Trans. JSASS Aerospace Tech. Japan Vol. 12, No. ists29 (2014)

Pk_8

2. Experimental Method

2.1. Wind tunnel Figure 1 shows “Planetary Atmosphere Wind Tunnel

(PAWT)” at JAXA/ISAS (Institute of Space and Astronautical Science). The specification is tabulated in Table 1. This tunnel is a low-density circuit wind tunnel. The driving system is a counter-rotating axial fan. The outer diameter of the propeller is 2.6 m. In the test section, the flow outlet diameter is 1.6 m and the flow inlet diameter is 1.9 m. The distance from the outlet to the inlet is 3.5 m. Two rotary vacuum pumps are connected to the wind tunnel and it takes approximately 30 minutes to reduce the pressure down to 1/10 atmosphere. Flow calibration tests were conducted in the previous study and the uniform flow region that is defined as a region where the flow velocity is within 1% dispersion from the mean flow velocity covers approximately 35% (area with a diameter of 550 mm) in the central part of the test section at 1/10 atmosphere.

A low-speed wind tunnel at KIT was used for a wind tunnel correlation and evaluation of tail airfoil performance. The cross section of the test section is 800 mm x 800 mm. The flow range is from 2.0 m/s to 13.0 m/s. The turbulence level is less than 0.3% when the flow velocity is 3 m/s. 2.2. Test Model

The test model is shown in Fig. 2 and its specification is tabulated in table 2. This model is a 1/5-scale model of a Mars airplane made of rapid prototyping resin. The airfoil shape of the main wing is Ishii airfoil with AR = 5.4. The main wing dividing into inner and outer wings has a dihedral angle of 5 deg. NACA0009 was used as both horizontal- and vertical-tail airfoils. The planform of the tail wing is a trapezoidal shape (Lower base: 90 mm, Upper base: 60 mm) and the distance between each vertical-tail wing is 60 mm. The angle of the

elevator can be changed by replacing the mounting plate with various elevator angles. The incidence angle of the tail is zero.

2.3. Measurement method and setup Figure 3 shows a body axis coordinate system (right-handed

coordinate system) defining positive direction for the force and moments acting on the body as well as its linear and angular velocity vectors at the location of a center of gravity of the body (cg). The cg is determined at the point located 30% chord length apart from the leading edge of the main wing.

In PAWT tests, a six-component sting force/moment balance (LMC-6402-M, NISSHO-ELECTRIC-WORKS) was used for the aerodynamic measurement. Aerodynamic force acting on the body is so small that the minimum drag falls below 1/100 of its rated capacity. Therefore, balance-system calibration tests were preliminarily conducted by loading in the actual measurement setup condition and then calibration curves and each interference coefficient were obtained. Meanwhile, in KIT tests, a three-component force/moment balance (LMC-3501-5N, NISSHO- ELECTRIC-WORKS) was used. The accuracy of the three-component balance is higher than that of the six-component balance in the PAWT because of its lower rated capacity.

Fig. 1. Planetary Atmosphere Wind Tunnel (PAWT).

Table 1 Specification of Planetary Atmosphere Wind Tunnel Tunnel Type low-density closed circuit

Test Section 1600 mm (Diameter)

Total Pressure 5 ~ 100 kPa

Maximum Flow velocity 1/5 atmosphere

163 m/s

Working Gas Air

Fig. 2. 1/5 Scale Model of Mars Airplane.

Table 2. Specification of 1/5 Scale Mode. Total length 400 mm Total width 515 mm

Main wing

Airfoil Ishii airfoil Chord length 95 mm

Area 48925 mm2

Dihedral angle 5 deg

Tail wing Airfoil NACA0009

Horizontal tail area 6283 mm2

Vertical tail area 2116 2 mm2

Fig. 3. Body Axis Coordinate System.

M. ANYOJI et al.: Wind Tunnel Tests on Aerodynamic Characteristics of a Mars Airplane Scale Model

Pk_9

Figure 4 shows an experimental setup at the PAWT. The test model was supported by the strut from below and installed in a direction perpendicular to the ground. The force balance was installed in the aft section of the strut and covered by a windshield. The angle of attack was varied by shifting the test model in a horizontal direction.

2.4. Experimental Condition The Reynolds number was changed from 10,000 to 33,000. The characteristics length of the Reynolds number was the chord length of the main wing of 95 mm. The wind velocity was kept at 27 m/s in all cases and the pressure inside the wind tunnel was changed in the range from 1/5 to 1/10 atmosphere when the Reynolds number effect was evaluated. Also, the angle of attack was changed from -20 deg to 20 deg. The pitching moment was evaluated at the elevator angle of 0 deg, -5 deg and -10 deg at Re = 33,000. The wind tunnel correlation test in the low-speed wind tunnel at KIT was carried out at only Re = 33,000.

3. Results and Discussion

3.1. Wind tunnel correlation A comparison between the PAWT results and the KIT one as a wind tunnel correlation at Re = 33,000 is shown in Fig. 5.

The lift-curve slope of the both results is in good agreement though there in a little difference in the maximum lift coefficient in high angles of attack. However, there is a large difference of the drag-coefficient level while both drag curves are similar in the wide angle of attack range. In particular, the

drag coefficient at the PAWT at the zero angle of attack (CD0)is approximately twice as large as that at the KIT. This difference in drag coefficient is due to a support interference effect. In the PAWT tests, the support drag correction was not applied yet and the results include its effect. By contrast, in

(a) Front side

(b) Top side Fig. 4. Experimental Setup at the PAWT .

-1

-0.5

0

0.5

1

-20 -10 0 10 20

JAXA/ISASKIT

CL

AoA [deg]

(a) Lift curve

0

0.1

0.2

0.3

0.4

0.5

-20 -10 0 10 20

JAXA/ISASKIT

C D

AoA [deg]

(b) Drag curve

-8

-6

-4

-2

0

2

4

6

8

-20 -10 0 10 20

JAXA/ISASKIT

L/D

AoA [deg]

(c) L/D Fig. 5. Wind tunnel correlation (Re = 33,000).

Trans. JSASS Aerospace Tech. Japan Vol. 12, No. ists29 (2014)

Pk_10

the KIT tests, the correction tests were performed and its effect was removed from the result. As a result, the drag at the PAWT becomes larger than that at the KIT. The effect of support drag at the PAWT is currently under investigation. The maximum lift-to-drag ratio (L/Dmax) at the PAWT becomes lower than that at the KIT that reaches 6.8 because of the difference in the drag coefficients. There is a remarkable difference in the drag coefficients and L/Dmax between both wind tunnel tests, however if the support drag effect is corrected, it is expected that the PAWT result approaches to the KIT result because the similar change in the drag with changing the angle of attack can be found. Therefore, it can be concluded that there is a correlation between the PAWT results and the KIT one. 3.2. Reynolds number effect on aerodynamic performance

Figure 6 shows the Reynolds number effect on the aerodynamic performance. Each lift curve changes little despite the change in the Reynolds number though there is a small non-linearity of the lift curve at = 4 deg at Re = 10,000. The lift-curve slope has a high linearity in a wide range of the angle of attack and the lift coefficient starts to level off at around = 10 deg where burst of a laminar separation occurs in the main wing 12. However, general stall characteristics with a drastic lift decrease that is seen in the much higher Reynolds number regime are not found in Fig. 6(a). Instead, the lift coefficient continues to increase up to the large angle of attack. At Re = 23,000 and 33,000, drag changes little while the drag at Re = 10,000 becomes larger as shown in Fig. 6(b) because the viscosity effect becomes pronounced. As these results, the L/Dmax reaches 4.5 at = 7 deg at Re = 33,000. On the other hand, the L/Dmax at Re = 10,000 reaches only 3.4 at most due to a larger drag than that at Re = 23,000 and 33,000. The moment coefficient about the center of gravity (Cmcg) has no Reynolds number dependence in the wide angle of attack range from = -5 deg to 12 deg as shown in Fig. 6(d). Furthermore, the longitudinal static stability derivatives ( Lcg CCm / ) become negative. This means that the longitudinal static stability was ensured. 3.3. Effectiveness of elevator The aerodynamic performance when the elevator angle ( e)is varied from 0 to -10 deg at Re = 33,000 are shown in Fig. 7. The lift and drag curves do not change in the positive angle of attack though there is a difference both in lift and drag curves in the negative angle of attack. The pitching moment characteristics are focused to evaluate the effectiveness of the elevator. It is found that there is little change in the moment curve between e = 0 deg and e = -5 deg and also the trim angle does not change from around 0 deg. In addition, the trim angle at e = -10 deg is approximately 3 deg. These results indicate that the effectiveness of the elevator has a non-linearity and the elevator has very little effect in low elevator-angle range less than -5 deg. Therefore, the effectiveness of the elevator is exceedingly small. The current design requires a large elevator-angle control even for a small pitching control. To improve the effectiveness of the elevator, a design change of the elevator is required.

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

-20 -15 -10 -5 0 5 10 15 20

Re = 33,000Re = 23,000Re = 10,000

CL,

CD

AoA [deg]

(a) Lift curve

-1

-0.5

0

0.5

1

0 0.1 0.2 0.3 0.4 0.5

Re =33,000Re =23,000Re =10,000

CL

CD

(b) Drag polar

-4

-2

0

2

4

-20 -15 -10 -5 0 5 10 15 20

Re = 33,000Re = 23,000Re = 10,000

L/D

AoA [deg]

(c) L/D

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

-20 -15 -10 -5 0 5 10 15 20

Re = 33,000Re = 23,000Re = 10,000

Cm

cg

AoA [deg](d) Pitching moment

Fig. 6. Reynolds number dependency on aerodynamic performance

M. ANYOJI et al.: Wind Tunnel Tests on Aerodynamic Characteristics of a Mars Airplane Scale Model

Pk_11

3.4. Design guideline for improvement of elevator effectiveness

It can be considered as reasons for the low effectiveness of the elevator that the interference of the vertical tail or the fuselage wakes. As for the former, wind tunnel tests that a distance between each vertical tail was expanded from 60 mm to 100 mm were conducted. However, it does not affect the effectiveness. To investigate the interference of the fuselage wakes, the tail was lifted up so that the horizontal tail is located above the fuselage. However, this modification also does not affect the effectiveness. As a result of a variety of tail design changes, the most effective factor against the pitching moment is found to be an airfoil shape of the horizontal tail. Figures 8 and 9 show pitching moments of the airplane whose tail airfoils are a thin flat-plate with the thickness of 1.3% of the chord length and NACA0006, respectively. Then,

the planform of the horizontal tail is not changed. In the 1.3% flat plate shown in Fig. 8, the moment curve at e = 0 deg is changed from that of the original design (the original tail airfoil is NACA0009) due to a change of the horizontal-tail airfoil. Compared to the results of the original design shown in Fig. 7 (c), the trim angle significantly varies even at the low elevator angle. At e = -5 deg and -10 deg, the variations of the trim angle from at e = 0 deg are 3.5 deg and 6 deg, respectively. Also, in NACA0006, the variations of the trim angle from at e = 0 deg are 4.2 deg at e = -5 deg and 7.8 deg at e = -10 deg, respectively. Although the variation of the trim angle of NACA0006 is larger than that of the 1.3% flat plate, the variation of Cmcg in NACA0006 when the e is changed is slightly smaller than that of the 1.3% flat plate. As these results, both the 1.3% flat plate and NACA0006 which are thinner tail airfoils than NACA0009 can improve the effectiveness of the elevator.

Figure 10 shows airfoil aerodynamic characteristics of NACA0012, NACA0009, NACA0006 and a 1.3% flat plate at Re = 20,000 corresponding to the vicinity of the Reynolds number based on the tail chord length in the case of Re = 33,000 based on the main-wing chord length. In such a low Reynolds number, it is known that a thinner airfoil has higher lift coefficients.15 In thick airfoils such as NACA0012 and NACA0009, non-linear increase in lift is found around at AoA = 4 deg and the lift-curve slope up to this angle of attack is much smaller than those of thinner airfoils such as NACA0006 and 1.3% flat plate.

The moment coefficient about cg at the zero-lift angle of attack of the main wing is expressed by the following equation:

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-20 -15 -10 -5 0 5 10 15 20

e = 0 deg

e = -5 deg

e = -10 deg

CL,

CD

AoA [deg]

(a) Lift and Drag

-1

-0.5

0

0.5

1

0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45

e = 0 dege = -5 dege = -10 deg

CL

CD

(b) Drag polar

-0.6

-0.4

-0.2

0

0.2

0.4

-20 -15 -10 -5 0 5 10 15 20

e = 0 dege = -5 dege = -10 deg

Cm

cg

AoA [deg]

(c) Pitching Moment

Fig. 7. Effectiveness of Elevator (e= 0 - -10deg, Re = 33,000).

-0.6

-0.4

-0.2

0

0.2

0.4

-10 -5 0 5 10 15 20

e = 0 dege = -5 dege = -10 deg

Cm

AoA [deg]

Fig. 8. Pitching Moment (Tail airfoil: 1.3% flat plate, Re = 33,000).

-0.6

-0.4

-0.2

0

0.2

0.4

-10 -5 0 5 10 15 20

e = 0 dege = -5 dege = -10 deg

Cm

cg

AoA [deg]Fig. 9. Pitching Moment (Tail airfoil: NACA0006, Re = 33,000) .

Trans. JSASS Aerospace Tech. Japan Vol. 12, No. ists29 (2014)

Pk_12

tailwingaccg LtLwmm ClClCC (1)

The moment coefficient about the main wing’s aerodynamic center (Cmac), the lift coefficient of the main wing (CLwing) and the distance from cg to ac of the main wing (lw)and the tail (lt) are constant regardless of the tail airfoil shape. Hence, only the lift coefficient of the tail (CLtail) affects Cmcg

when the tail airfoil is changed. If a variation of the zero-lift angle of attack in each tail

airfoil shown in Fig. 10 is equal when the elevator angle is changed, thinner airfoils with larger lift-curve slope are more effective for the pitching control of the airplane because thin airfoils have higher lift coefficients than those of thick airfoils compared with each lift curve at the same angle of attack, especially in low angles of attack. Hence, the lift-curve slope and its linearity of the tail airfoil is a dominant factor for the pitching control. From this perspective, the non-linear variation of the lift-curve slope with the elevator angle is thought to be the cause of the non-linear effectiveness of the elevator shown in Fig. 7(c). It is considered that the lift coefficient of the tail airfoil (NACA0009) at e = -5 deg is not as large as it affects the pitching moment of the airplane. However, judging from the variation of the trim angle with the elevator angle shown in Fig. 7(c) and Fig. 9, the lift-curve slope and the lift curve itself of NACA0009 at e = -10 deg is estimated as large as the lift coefficient of NACA0006 at e =-5 deg. The aerodynamic performance of the tail airfoil with control surface will be also investigated as future works. Eventually, the validation of the horizontal tail lift-slope coefficient and the tail volume is evaluated by a flight simulation.

4. Conclusion

In this research, to evaluate the aerodynamic performance of a Mars airplane at low Reynolds numbers (Re =10,000-33,000), low-density wind tunnel tests were carried out using a 1/5-scale model. Reynolds number dependency on the lift curve and the pitching moment is found to be little. However, depending on the tail airfoil, the effectiveness of the elevator was low for responsive pitch control. The low effectiveness of the elevator was due to low lift-curve slope and a non-linearity of the lift curve of the horizontal-tail

airfoil. A thinner tail airfoil with linear and large lift-curve slope is highly effective. Therefore, the aerodynamic performance of tail airfoil strongly affects the effectiveness of the elevator in such a low Reynolds number region. Improvement of the effectiveness of the elevator is expected by changing the current tail airfoil with NACA0006 or a thin flat plate though the thickness of the tail for the servo system installation should be also considered. The lateral stability and the effectiveness of the other control surface such as aileron and rudder have been evaluated.

Acknowledgments

We would like to express our sincere thanks to Hiroki Nagai at Tohoku University and Koichi Yonemoto at Kyushu Institute of Technology. This research was partially supported by KAKENHI (23860076) and (24246136).

References 1) Guynn, M. D., et al: Evolution of a Mars Airplane Concept for the

ARES Mars Scout Mission, 2nd AIAA “Unmanned Unlimited” conference and Workshop & Ehibit, 2003, 2003-6578.

2) Robert D. Braun and David A. Spencer: Design of the ARES Mars Airplane and Mission Architecture, Journal of Spacecraft and Rockets, 43 No.5 (2006), pp. 1026-1034.

3) K. Rinoie: Mars Airplane for Geographical Exploration -Conceptual Design Results from Student Design Projects, The 2nd International Symposium on Innovative Aerial/Space Flyer Systems, Univ. of Tokyo, 2005.

4) A. Oyama, K. Fuji: A Study on Airfoil Design for Future Mars Airplane, AIAA-2006-1484, 44th AIAA Aerospace Sciences Meeting and Exhibit , 2006, 2006-1484.

5) P. B. S. Lissaman: Low-reynolds-number airfoil, Annual Review of Mechanics, 15 (1983), pp. 223-239.

6) M.M. O’Meara and T.J. Mueller: Laminar Separation Bubble Characteristics on an Airfoil at Low Reynolds Number, AIAA Journal, 25 No.8 (1987), pp. 1033-1041.

7) A. Azuma, et al: Aerodynamic Characteristics of Wing at Low Reynolds Number, Fixed and Flapping Wing Aerodynamics for Micro Air Vehicle Applications, edited by T. J. Mueller, Progress in Astronautics and Aeronautics, Vol. 195, AIAA, Reston, VA (2001), pp. 341-398.

8) T. J. Mueller and S. M. Batill: Experimental Studies of Separation on a Two-Dimensional Airfoil at Low Reynolds Number, AIAA Journal, 20 No.4(1982), pp. 457-463.

9) E.V. Laitone: Aerodynamic Lift at Reynolds Number Below 7 x 10 exp 4, AIAA Journal, 34 No.9 (1996), pp. 1941-1942.

10) Schmitz, F. W.: Aerodyamics of the Model Airplane Part1, RSIC-721, 1967.

11) Schmitz, F. W.: The Aerodyamics of Small Reynolds Number, NASA TM-51, 1980.

12) M. Anyoji, et al: Aerodynamic Characteristics of Ishii Airfoil at Low Reynolds Numbers, Eighth International Conference on Flow Dynamics, 2011, OS6-11.

13) K. Nose, et al: PSP Measurement of Pressure Distributions on a Cambered Airfoil for Mars Exploration Airplane, The 11th Asian Symposium on Visualization, 2011, ASV11-11-10.

14) A. Oyama: Multiobjective Design Exploration of Airplane for Mars Exploration, 21st Workshop on Astrodynamics and Flight Mechanics, 2011.

15) M. Okamoto, K. Yasuda and A. Azuma: Aerodynamic Characteristics of the Wings and Body of a Dragonfly, J. Exp. Biology, 199 (1996), pp. 281-294.

0

0.2

0.4

0.6

0.8

1

0 5 10 15 20

NACA0012NACA0009NACA00061.3% flat plate

C L

AoA [deg]

Fig. 10. Aerodynamic performance of tail airfoils (Re = 20,000 [based on tail chord length]).